1. Field of the Invention
The present invention relates to a document retrieving system and a document retrieving apparatus for retrieving a document from documents registered in two or more document management servers which are connected via a network, and method, program and a storage media therefor.
2. Description of the Related Art
A document retrieving system is known in which two or more documents are stored in a database such as a document server, and a user retrieves a desired document from the stored documents. In such a system, when a document is registered for example, keywords contained in the document data are extracted to produce an index. The index produced in this manner is managed separately from the document in association with that document.
When a user inputs a keyword for retrieving a document, it is determined if the input keyword is included in the indices. If an index includes the keyword, the document corresponding to that index is determined as the document to be retrieved. Using the index in this manner improves the response time during retrieval.
As such a document retrieving system, implementation of an integrated document retrieval service is proposed by an apparatus that the user operates (personal computer (PC), for example) requesting two or more servers for retrieval (see Japanese Patent Laid-Open No. 2004-342042). In such a system, two or more servers share the indices of the documents stored in the respective servers. Accordingly, by the user issuing a document retrieval order to any one of the servers, it is possible to perform retrieval from the documents stored in two or more servers.
However, the above-described conventional technique has the following problems. That is, upon retrieving a document from documents stored in two or more servers or the like that are connected via a network, if a retrieval request is made to all the servers, it may take time before receiving retrieval results from all the servers. In addition, since a large volume of data containing information on the retrieval request or retrieval results is sent to the network for each retrieval operation, the network will be under heavy load.
Alternatively, if two or more servers share the indices of the documents stored in the respective servers, as disclosed in Japanese Patent Laid-Open No. 2004-342042 described above, it is sufficient to make a retrieval request to one of the servers upon retrieval, which improves the retrieval efficiency.
However, in such a case, two or more servers are respectively required to maintain the indices of the documents registered in all the other servers. Therefore, as the number of servers or stored documents increases, the data volume of the indices to be maintained increases, which invites a waste of memory resources and extends the time required for retrieval. Also, in a case that a document is registered in a certain server, the index thereof is transferred to all the other servers via a network, which increases network traffic.